In a move that underscores the escalating tension between the U.S. and China over technology, Intel has informed Chinese clients it will now require licenses to export certain high-performance artificial intelligence (AI) chips. The development, first reported by the Financial Times, follows fresh trade restrictions imposed by the U.S. government, further tightening the grip on advanced semiconductor exports to China.
This shift places new hurdles in front of American chipmakers like Intel and Nvidia, both of which have seen significant demand for AI chips from China despite mounting geopolitical risks.
According to sources cited in the report, Intel notified Chinese clients last week that U.S. regulators will now require a license to export any chip that meets or exceeds the following thresholds:
- DRAM bandwidth of 1,400 GB/s or more
- Input/Output (I/O) bandwidth of 1,100 GB/s or more
- Or a combined bandwidth of 1,700 GB/s or higher
These thresholds are specifically targeted at curbing China’s access to cutting-edge AI chips, which are essential for training large language models, computer vision systems, and other advanced AI capabilities. Intel’s Gaudi series—designed for AI workloads—reportedly exceeds these limits, meaning the company must now apply for licenses before shipping such chips to clients in China.
Parallel Impact on Nvidia and ASML
The announcement came just one day after Nvidia issued a warning that it expects a $5.5 billion revenue hit as a direct result of these restrictions. Nvidia’s H20 chip, developed specifically to comply with earlier U.S. export controls, also now falls under the new rules, effectively slamming the brakes on a workaround the company had hoped would maintain its market share in China.
Meanwhile, Dutch semiconductor equipment manufacturer ASML also signaled concern, saying export uncertainties have clouded its business outlook. These developments point to a broader ripple effect across the global semiconductor supply chain, hitting chipmakers and equipment manufacturers alike.
Geopolitical Chess Match Intensifies
The latest move is part of a wider U.S. effort to contain China’s technological rise, particularly in areas deemed crucial for national security, such as artificial intelligence, quantum computing, and next-generation telecommunications.
Under the Biden administration, export controls—first initiated during Donald Trump’s presidency—have not only persisted but expanded. These rules are intended to prevent Chinese companies, particularly those linked to the military or surveillance infrastructure, from acquiring state-of-the-art semiconductors that could enable rapid technological advancement.
Intel’s response illustrates the challenges faced by global firms attempting to navigate a landscape where commercial and political priorities are increasingly in conflict.
Intel’s Position and Market Reaction
Intel has not issued a public comment in response to the report. However, the market reaction was swift. The company’s shares fell over 3% in Wednesday’s trading, mirroring broader declines in the semiconductor sector. Investors are becoming increasingly wary of the sector’s exposure to political risk, especially as the U.S. ramps up enforcement of export controls.
Intel CEO Lip-Bu Tan, who took over leadership at a critical moment, now faces the dual task of steering the company through an AI-driven market boom while complying with increasingly restrictive trade policies.
The End of the AI Chip Boom?
Over the past two years, AI chipmakers like Nvidia and Intel experienced a massive rally, driven by surging demand for generative AI and large-scale computing. However, that growth is showing signs of slowing, as governments intervene and corporate buyers grow cautious amid rising costs and regulatory uncertainty.
Tariff threats, export bans, and new compliance requirements are weighing heavily on investor sentiment. There’s also growing concern that “Big Tech” spending on AI infrastructure could decelerate, particularly if economic conditions tighten in 2025.
For China, the writing has been on the wall for some time. The country has invested billions into developing domestic semiconductor capabilities, in an attempt to reduce reliance on U.S. technology. But progress has been slow—especially in the high-performance AI segment, where U.S. firms still maintain a significant lead.
With restrictions tightening, Chinese companies may be forced to look toward less powerful alternatives, increase their use of open-source hardware, or ramp up R&D in hopes of creating competitive in-house chips. But even in a best-case scenario, bridging the technological gap will take years.
As the global chip war deepens, Intel’s licensing requirement marks yet another escalation in a long-term battle over technological dominance. While the U.S. hopes to slow China’s AI progress, the move could have unintended consequences, including disrupting global innovation and fragmenting the tech supply chain.
For Intel and its peers, the challenge is to balance compliance with competitiveness, continuing to lead in AI development while navigating a fractured and highly politicized market. Whether this latest policy accelerates the global race toward AI self-sufficiency—or just stalls it temporarily—remains to be seen.
